v1.6.0

The specific versions of each component can be found in the README.

Updates per WG:

• AutoML WG
• Notebooks WG
• Pipelines WG
• Training WG
• KServe
• KFP Tekton

v1.6.0-rc.3

Fixes included in the RC:

• Notebook invalid kustomization patch introduced for profile-controller manager kubeflow/kubeflow#6604

v1.6.0-rc.2

Fixes included in the RC:

• Notebook Missing image groups
• Notebook Profiler annotation
• Notebook CRD missing status field
• Notebook Conditions being appended
• Notebook Endless restarts
• Katib fix for pytorch-mnist image issue

v1.6.0-rc.1

Fixes included in the RC

• Fix typo in the networkpolicies #2254
• Sync kserve web app manifests v0.8.1 #2256
• Update kubeflow/kubeflow manifests from v1.6.0-rc.1 #2260
  • Support k8s 1.22 in profile controller kubeflow/kubeflow#6491
  • Reload certificates in admission-webhook kubeflow/kubeflow#6581

v1.6.0-rc.0

This is the first RC of the Kubeflow 1.6 release 🎉

The next phases of the release will include updating the docs, manifest testing, and distribution testing. For a more detailed overview of the schedule take a look at the 1.6 release timeline

v1.5.1

Includes the for KFP cache-deployer and the cert generation mechanism #2165. This affects K8s clusters with more strict security policies, in which users can't create certificates with kubelet's signer name.

Special thanks to @akartsky and @surajkota for driving this effort!

v1.5.0

The specific versions of each component can be found in the README.

Updates per WG:

• AutoML WG
• Notebooks WG
• Pipelines WG
• Training WG
• KServe

v1.4.1

Patch release to include the fix for #2082

v1.4.0

The specific versions of each component can be found in https://github.com/kubeflow/manifests/tree/v1.4.0#kubeflow-components-versions.

Updated per WG:

• AutoML WG
• Notebooks WG
• Pipelines WG
• Serving WG
• Training WG

v1.4.0-rc.2

Include the latest Training Operator v1.3.0-rc.2 #2032

Includes fixes for:

• GoogleCloudPlatform/kubeflow-distribution#317